The Science Journal of the American Association for Respiratory Care

2010 OPEN FORUM Abstracts

AUTO-PEEP DURING APRV VARIES WITH THE VENTILATOR MODEL

Ehab Daoud, Robert L. Chatburn; Critical Care, Cleveland Clinic, Cleveland, OH

BACKGROUND: Airway pressure release ventilation (APRV) is a form of the “open lung approach” ventilation strategy used for acute respiratory distress syndrome (ARDS) and it is incorporated on most modern ventilators though under different names. This mode is used usually with very short release time (T Low) to create auto-PEEP and prevent alveolar collapse at end of the T Low. No data are available to estimate the amount of auto-PEEP in APRV using varying T Low. No studies have been done to compare ventilator performance in regards to auto-PEEP and flow delivery. The purpose of this study was to compare auto-PEEP levels, peak expiratory flows (PEF) and flow decay profiles among 4 common intensive care ventilators. METHODS: A passive ARDS patient was modeled with a lung simulator (IngMar ASL 5000): compliance = 30 mL/cm H2O, resistance = 10 cm H2O/L/s, with time constant (TC) of 0.3 seconds. Ventilator settings: P High (25 cm H2O), P Low (0 cm H20). T Low settings were varied from 1 to 6 TC, i.e. from 0.3 seconds to 1.8 seconds. T High was held constant. We measured auto-PEEP and flow as %PEF at each TC, where time was measured from the start of expiratory flow. Delay was defined as the time between start of expiratory flow and PEF. System TC was defined as the time for flow decay from PEF to 37% of PEF. Four ventilators were evaluated: Carefusion Avea, Draeger Infinity, Covidien PB 840, Maquet Servo I. Mean values from at least 5 breaths were compared with 1 way ANOVA on ranks. RESULTS: Mean data are summarized in Table below. At all T Low the Servo-i had the lowest and Avea the highest auto-PEEP. CONCLUSIONS: All tested ventilators did not follow the mathematical model as expected and resulted in significantly different auto-PEEP, flow decay, and PEF regardless of the T Low settings. An unexpected finding was that peak flow was delayed from the start of expiratory flow by 0.11 to 0.17 ms, which affected the flow decay data. System TC was larger than model TC because it included both model and ventilator circuit resistance and compliance and was affected by the delay. Setting T Low according to predicted time constants may not be reliable and results in higher than expected auto-PEEP. Sponsored Research - None